Potential Implications of Sea-Level Rise for Belgium

Journal of Coastal Research 24 2 358–366 West Palm Beach, Florida March 2008

Potential Implications of Sea-Level Rise for BelgiumLuc Lebbe†, Nathalie Van Meir†, and Peter Viaene‡

†Ghent UniversityDepartment of Geology and Soil ScienceLaboratory of Applied Geology and

HydrogeologyKrijgslaan, 281 (S8)B-9000 [email protected]

‡Ministry of the Flemish CommunityEnvironment and Infrastructure

DepartmentWaterways and Maritime Affairs

AdministrationHydraulic Laboratory and Hydrological

Research DivisionBerchemlei 115B-2140 BorgerhoutBelgium

ABSTRACT

LEBBE, L.; VAN MEIR, N., and VIAENE, P., 2008. Potential implications of sea-level rise for Belgium. Journal ofCoastal Research, 24(2), 358–366. West Palm Beach (Florida), ISSN 0749-0208.

The Belgian coastal plain and the Schelde estuary are threatened by sea-level rise. While of great economic importancewith a threatened population of some 0.8 million (of a total population of 10 million), assessments of these risks arelimited. This article describes the physical characteristics of the coast and undertakes a qualitative interpretation ofits vulnerability. Low-lying polders are the most vulnerable to sea-level rise where a major problem is water drainageduring rainy periods; their varying vulnerability to sea-level rise and increase in rain intensity is assessed, includingthe relationship between drainage levels and saltwater seepage. Freshwater lenses developed within the dunes arealso vulnerable to sea-level rise, leading to threats to drinking water supplies from saltwater intrusion. Belgian coastaldefence structures and their effectiveness are discussed. Historical sea-level rise during the past century, wave andwind data, and the evolution of erosion and accretion along the coast are interpreted. For Antwerpen, a harbour cityon the river Schelde, the effects of sea-level rise are far from clear. Included here are historical data on changes intidal amplitude during the 20th century. Future research needs should focus on the quantitative interpretation ofdata to understand the effect of sea-level rise on beach erosion, flood risk, and fresh and salt groundwater distribution.Furthermore, a thorough socio-economic study should be undertaken to assess the vulnerability of the Belgian coastand the Schelde estuary.

ADDITIONAL INDEX WORDS: Belgian coastal plain, saltwater intrusion, gravitational and artificial water drainage,coastal defences.

PHYSICAL CHARACTERISTICS OF THE BELGIANCOASTAL PLAIN

Geographical Position and Economic Relevance

The Belgian coastline is on the southern North Sea andcomprises an open wave-exposed coast and the more shel-tered upper reaches of the Schelde estuary. The wave-ex-posed coast (65-km long) has a nearly continuous dune beltthat protects the hinterland. This dune barrier consists of twoarcs, one stretching from Dunkerque (France) to Wenduineand a second arc from Wenduine to Breskens (The Nether-lands). Its continuity is interrupted by the estuaries of theriver IJzer in the west and het Zwin in the east and the har-bours of Oostende, Blankenberge, and Zeebrugge. The totalcoastal zone population is ca. 0.4 million (i.e., ca. 4% of theBelgian population), which is concentrated in the towns ofBrugge, Oostende, Blankenberge, Veurne, Nieuwpoort, andKnokke (Figure 1).

During summer holidays the population increases by ca.0.3 million resident tourists and on peak days by 0.25 million

DOI: 10.2112/07A-0009.1 received and accepted in revision 19 April2007.

day trippers. The economic importance of the coast is mainlya consequence of beach tourism and agriculture in the low-lying polder areas with the combined economic activities fromtourism, fishing, and harbour activity generating €2.7 billionper year (ALLAERT, 1996).

Although the Schelde estuary is mainly situated in TheNetherlands, its socio-economic importance for Belgium isparamount. The harbour city of Antwerpen (population, 0.4million) is situated in an area of low-lying polders in the up-per reaches of the Schelde estuary (Figure 2). Two Dutch har-bours, Vlissingen and Terneuzen, are located along the Schel-de estuary, whereas another Belgian harbour, Gent, is con-nected to the estuary by a sea canal. Therefore, a cooperationexists between The Netherlands and Flanders (the northernpart of the Federal State of Belgium) on harbour policy de-cisions. Of these four harbours, Antwerpen is the most im-portant economically. The yearly added values of harbour ac-tivities (1997/1998 average) are €9.1 billion (Antwerpen),€3.1 billion (Gent), €2.1 billion (Vlissingen), and €1.5 billion(Terneuzen). The harbours of Antwerpen and Gent providejobs for 153,000 people directly and indirectly, and their so-cio-economic importance for Belgium is significant. The es-tuary is also used for sand extraction, fisheries, and nature

359Sea-Level Rise in Belgium

Journal of Coastal Research, Vol. 24, No. 2, 2008

Figure 1. The Belgian coastal plain, its geographical position in the NorthSea, the landward limit of 5 m TAW, indication of dune areas, and thelocation of the resistivity profile R-R�.

Figure 2. The Schelde estuary and the indication of the harbour cities An-twerpen, Gent, Terneuzen, and Vlissingen (adapted from Baeteman et al.,1992).

reserves. The morphology of the Schelde estuary has changedsignificantly due to natural and human actions, which hashad repercussions for the functionality of the area. Humanaction in recent years has consisted of dredging, changes inchannel width, and construction of river defences. It is ex-pected that tides and the river bed will have adjusted to theimpact of these activities in 10 to 15 years (ZANTING and TEN

THIJ, 2001), assuming there is no further interference.

Geology

The clastic sediments of the coastal plain are of Quaternaryage and are underlain by a thick Tertiary clay formation(MVG, 1993; VIAENE, 2000). They form a wedge that is thick-est in the east, thins to the west, and also decreases from thecoast in the direction of the polders. Most of these sedimentswere deposited after the last Glacial Stage some 10,000 yearsBP, mainly after sea-level rise slowed down some 6000 yearsBP, and are of Atlantic age or younger. Pleistocene sedimentsare also present, but to a lesser extent (DENYS et al., 1983;DENYS and BAETEMAN, 1995; MOSTAERT, 1985). The Qua-ternary sediments consist of sand, coarse sand with shells,clay, and peat layers. The reduction in sea-level rise about6000 years BP was also responsible for the form of the coastalplain: a natural barrier of dunes, shore, and foreshore withno tidal inlets (a closed coast) protecting a low-lying plain.

Geomorphology and Occupation

The landward limit of the Belgian coastal plain is definedby the topographical contour of 5 m TAW1 (Figure 1). Be-tween the dunes and hinterland a landscape characterised bychannels and intertidal flats formerly existed, but is now oc-cupied by polders. Despite the presence of a dune barrier, the

1 The Belgian datum level (TAW, Tweede Algemene Waterpassing,Second General Leveling) corresponds with the low low water atOostende (1839–1858); 0 m TAW is 2.33 m below the zero of NAP(Datum Level of the Netherlands), which corresponds to averagemean sea level during the last 300 years.

sea has entered the hinterland during storms at severalbreaching points and through tidal channels. From the 11thcentury onward people inhabiting the coastal plain startedprotecting their land and houses by building dikes and cre-ating polders. This process has shaped the present coastalplain with sand-filled channels and creeks producing higherland levels than the mudflats consisting of clay and peat,which have experienced settlement and consolidation (MOS-TAERT, 1986; MVG, 1993). The plain is now subdivided inthree parts: the low-lying polder areas, the dunes, and theshore with the low-lying polder areas forming the major partof the coastal plain. The width of the low-lying coastal plainranges between 10 km and 18 km and can be subdivided intofour characteristic regions: the Old Land, the Middle Land,the New Land, and the Deep Polders (TAVERNIER et al.,1970). The Old Land was brought under cultivation duringthe eighth and the ninth centuries and has elevations be-tween 2.3 m TAW and 3.8 m TAW. It is located behind thefirst line of dikes, the alignment of which is perpendicular tothe coastline protecting the inhabitants from flooding fromtidal channels. In the Old Land there is a pronounced inver-sion of the landscape due to differential settling of clay-cov-ered peat areas and sandy creeks, with the creeks becomingthe higher areas. Farmhouses are mostly situated on sandycreek ridges. The Middle Land was reclaimed during the 11thand 12th centuries and is situated on former tidal channels.Due to a thicker clay cover throughout this area the land-scape inversion is less pronounced and human occupation ismore evenly distributed than in the Old Land. The New Landwas reclaimed after the 12th century and has a limited lat-eral extension. These rather small areas are very flat, andthe ground level is situated above 4 m TAW. The Deep Pol-ders are the two lowest areas comprising lake beds that werereclaimed during the 17th century. Their extent is rather lim-ited (ca. 2300 ha) with ground level lying between 1 m TAWand 2 m TAW.

The coastal dunes range in height between 5 m TAW and30 m TAW, with the majority between 7 m TAW and 15 mTAW contributing substantially to the defence of the polders.

360 Lebbe, Van Meir, and Viaene


At the western Belgian coastal plain (west of the IJzer es-tuary) the dunes are approximately 2 km wide. East of theIJzer estuary the dune belt is only a few hundred metreswide, except for three places (Westende, De Haan, and Knok-ke) where the maximum dune width is about 2 km. The fu-ture natural development of the dune belt is inhibited by thepresence of large urbanised areas and a four-lane coastalroad. Nature reserves cover a large area in the dunes of thewestern coastal plain (738 ha). The largest reserve, De Wes-thoek, contains almost every representative type of dune veg-etation and forms part of a cross-border reserve along withthe French reserves La Dune du Perroquet and La Dune fos-sile de Ghyvelde. Two other important reserves on the westcoast are the Houtsaegerduinen and the Ter Yde dune com-plex. East of the IJzer estuary nature reserves in the dunesare strongly fragmented with the exception of the D’Heyedunes of Bredene and De Haan.

The foreshore, backshore, and the dunes constitute a nat-ural defence against the sea. The shape and position of thecoastal barrier is heavily dependent on the direction andstrength of the winds, tides, currents, and waves. Waves aremainly driven by westerly winds and by periodical stormsurges with a southwest, west, or northwest wind direction.Tides have an amplitude of 2.6–5.4 m and are semidiurnaltides. The major part of the Belgian intertidal zone has agentle slope (approximately 1%), and the distance betweenlow and high water springs ranges from 100 m to 600 m. Thesediments on the shore are almost exclusively well sorted fineto medium sands (median 0.18 mm). There are peat outcropson the foreshore from 3–5 km west of Oostende. The onlynature reserve on the shore is situated east of the Zeebruggeharbour dike. The nature reserves of the IJzer Estuary andhet Zwin include mud flats, saltmarshes, low-lying polders,dunes, and shore. The area around the harbour also consistsof low-lying polder areas.

Hydrogeology

Under rising sea levels, salinisation is a major risk. Thesalt–freshwater interface within the groundwater of the Bel-gian coastal plain was first mapped by a geo-electrical survey(DE BREUK et al., 1974). A detailed distribution of fresh,brackish, and saltwater was determined throughout thewestern Belgian coastal plain by performing a number of re-sistivity borehole cross-sections (LEBBE and PEDE, 1986) (seee.g., Figure 3). In the polders, the phreatic aquifer consistsmainly of pervious sediments in which a large freshwaterlens occurs above saltwater.2 Intensive superficial drainagein these areas obstructs the deep infiltration of freshwater.In the Deep Polders, brackish water is found due to the mix-ing of infiltrating freshwater and an upward flow of salinegroundwater.

In the dune areas, the phreatic aquifer is completely filledwith freshwater (Figure 3) that is exploited for water supplyto its maximum capacity (e.g., at Koksijde and Knokke). AtDe Panne there is overexploitation and the freshwater lens

2 The saltwater is connate, i.e., trapped in the sediments sincedeposition.

is threatened by saltwater flowing seaward from the polderstoward the dunes. Salinisation from the sea is also a threatto the aquifer in the dunes because on the higher part of theshore saltwater infiltrates during high tides and during lowtides there is seepage of saltwater on the lower part of theshore. This shallow saltwater cycle results in a saltwater lensin the upper part of the unconfined aquifer along the shoreabove the freshwater. This salt–freshwater inversion is typ-ical for shores with a pervious subsoil, a rather low inclina-tion, and fairly high tidal fluctuation and/or quite high waves(LEBBE, 1999). The water-collection areas in the dunes arenot threatened by sea-level rise, as can be deduced from thegeneral saltwater intrusion literature (e.g., BEAR, 1979; DO-MENICO and SCHWARTZ, 1990; FREEZE and CHERRY, 1979;TODD, 1980). However, higher sea levels do increase salt-water infiltration and may increase salinity in the water col-lection areas as well as promote landward seepage of salt-water into the polders behind the narrow dunes. In the nar-row dunes the freshwater lens sits above saltwater, whichflows very slowly landward into the low-lying polders. Thissalinisation process is already a problem that sea-level risewill exacerbate.

COASTAL CHANGE

This section considers coastal change and its drivers alongthe Belgian coast, with an emphasis on sea-level change andcoastal monitoring.

Sea-Level Change

Sea level has been rising since the end of the Last GlacialStage some 10,000 years ago. Sea level was then some 100m lower than today. Between 15,000 and 7000 BP, sea levelrose rapidly. Subsequently, sea-level rise continued but at adecreasing rate on the order of 0.1–0.2 mm/y during the last1000 years. Sea level is still adjusting to the effect of the lastglaciation/deglaciation (WARRICK, 1993). DENYS and BAE-TEMAN (1995) and BAETEMAN and DENYS (1997) constructeda sea-level rise curve for the Belgian coast (although most oftheir basal peat cores used for carbon dating were derivedfrom the western coastal plain). They found two marked re-gressions: (i) between 7500 and 7000 calibrated years BP and(ii) around 5500 to 5000 calibrated years BP. In an attemptto investigate sea-level rise in the very recent past (the last2500 years), LOUWYE and DECLERCQ (1998) focused their at-tention on indicators such as height of tidal levees, height ofmature salt marshes, and height of upper tidal flat deposits.Their results indicate a slow rise of spring high water levelsfrom the 17th century to the present day. No evidence wasfound for relative sea-level rise between AD 600–800 and themedieval climate optimum between AD 1000 and 1350. Theauthors conclude, provisionally, that during the last 700years the local mean spring high water could have risen amaximum of 0.55 m, which is a considerably slower rate thanthat observed during the last 160 years.

DE KRAKER (1999) studied Flemish documents originatingGfrom a polder region between Gent and Antwerpen datingfrom the period 1488 to 1609. These documents, which arecontinuous and homogeneous, contain annual dike invest-



Fig

ure

3.F

resh

,br

acki

sh,

and

salt

wat

erdi

stri

buti

onin

acr

oss

sect

ion

perp

endi

cula

rto

the

coas

tli

ne(r

esis

tivi

typr

ofile

R-R

�,fo

rlo

cati

onse

eF

igur

e1)

(Fro

mL

ebbe

and

Ped

e,19

86).



Table 1. Variation in high water level (HW), mean sea level (MS), and lowwater level (LW) (Van Cauwenberge, 1999).

StationHW

(m/decade)MS

(m/decade)LW

(m/decade)

Oostende (1927–1998) 0.0217 0.0144 0.0084Zeebrugge (1964–1998) 0.0170 0.0150 0.0086Nieuwpoort (1967–1998) 0.0412 0.0277 0.0188

ments and record nearly every storm that caused damage tothe dikes. To assess the impact of storminess, eight separatestorm categories were distinguished based on three princi-ples: (i) the exact wording in the text, (ii) the circumstancesin which a storm appeared, and (iii) the consequences of astorm, e.g., area flooded. Interpreting the data using thismethod indicates a small increase in storminess during thefirst quarter of the 16th century and a large increase instorminess during the second half of the 20th century. Thestorm surge records of Vlissingen (1848–1990) containing thedate and duration of storm surges and the maximum levelfor each high tide of the storm surge reveal a correspondingincrease in storminess beginning around the 1950s.

In Belgium three tide gauges are installed along the coast-line, situated in Oostende, Zeebrugge, and Nieuwpoort. Mea-surements are taken and interpreted by the Coastal Hydro-graphic Service (part of the Waterways and Maritime AffairsAdministration), and the information is then passed on to thePermanent Service for Mean Sea Level. Oostende has thelongest available record, from 1835 to present. Unfortunately,the most reliable record begins in 1927, and its continuitywas interrupted during World War II. For Zeebrugge and Ni-euwpoort, data are available from 1964 and 1967 onward,respectively. A linear curve fitting gave the results listed inTable 1. The correlation coefficients are statistically signifi-cant and show a rising trend (Figure 4).

The higher values derived for the Nieuwpoort station arepartly due to a local subsidence of the quay wall. The otherresults are comparable to the results found for neighbouringcountries as well as the global mean sea-level change calcu-lated for the past century (15–20 cm) (e.g., DOUGLAS, 2001).The results support the assumption that Belgium is geologi-cally stable and not prone to significant regional subsidence,although localised subsidence does occur e.g., Nieuwport andin the polders.

It should be noted that the high water (HW) level is in-creasing more than the low water (LW) level, indicating anincrease in tidal amplitude (VAN CAUWENBERGE, 1999). Thisis supported by the data for Oostende and Vlissingen whenanalysed with cyclic curve fitting, which is based on a nodaltidal cycle of 18.61 years. The time series from Oostende has3.67 nodal cycles, and the time series from Vlissingen has5.86 nodal cycles. Cyclic curve fitting suggests an increase inHW at Oostende of 2 mm/y and an increase in LW of 1 mm/y.In Vlissingen the increase for HW is 3 mm/y, but again about1 mm/y for LW. In contrast to The Netherlands (Vlissingen),the Belgian coast is considered geologically stable, so regionalland subsidence is assumed to be negligibly small. For Ni-euwpoort and Zeebrugge the time series of tide gauge obser-vations is too short for this type of regression analysis.

Tide gauge data during the past 100 years also exist forthe river Schelde in Antwerpen. This also indicates an in-crease in tidal amplitude as mean HW levels rise and meanLW levels decrease (Table 2). The 56-cm rise in high waterlevels during a period of 90 years is obvious and has raisedflood risk along the estuary. The reason for the increase ofthe mean HW level at Antwerpen is mainly due to humanmodifications in and around the estuary and partly due torelative sea-level rise. Human modifications comprise dredg-ing, dike construction, urbanisation, rectifying the course ofthe estuary, etc. The lowering of the mean LW level in An-twerpen is probably due to an improved low tide flow in theWesterschelde and the lower Zeeschelde (VIAENE, 2000). Theheight of extreme storm floods is a basic criterion for dikeconstruction, and an increase in storm flood heights are ob-served. This increase is due to a rising level of high tides, anda changing definition of ‘‘storm flood height’’ (i.e., �6.5 mTAW before 1980 and �6.6 m TAW after 1980). However,this does not imply a significant increase in the incidence offloods, as the dikes around Antwerpen have already been, orare being, raised to protect against an extreme water fre-quency of 1/4000 years (ZANTING and TEN THIJ, 2001).

Erosion and Accretion

Erosion and accretion along the Belgian coast are carefullymonitored. Since 1983 an annual aerial survey of the fore-shore, backshore, and dunes has been carried out in springover the entire coast with additional flights in autumn overthe most threatened parts of the coast. Since 1985 more in-formation has been gathered with a hydrographic hovercraftfor the acquisition of bathymetric nearshore data, which haverevealed a sequence of alternating erosional and accretionalsections (DE WOLF et al., 1993). The most marked example isaround De Haan where coastal defence structures do notchange the course of an erosive megaprotuberance. DE MOOR

(1992) defines this as

a longer term coastal dynamics component, embracing acoastal section of a few kilometres length and character-ised by a residual erosion continuing over several de-cades, changing in intensity and sweeping along thecoastline slowly in the direction of the residual current,so that it locally presents a longer term cyclic characterand is followed by an accretional phase.

This has led to a significant landward retreat of the coastlinenear De Haan (ca. 70 m at the low waterline from 1983 to1992). Despite the application of the Longard system andbeach nourishments, sand losses of almost 20 m3/m/y are ob-served (DE WOLF et al., 1993). Most accretion along the coast-line is related to defence structures, e.g., accretion at Koksijdeand Nieuwpoort can be attributed to new groynes, and nearWenduine and Blankenberge to new groynes, beach scrap-ings, and nourishment. The construction of osier hedgesalong the coastline has trapped large quantities of sand since1983 (DE WOLF et al., 1993). Without these numerous protec-tion measures, natural evolution along the Belgian coastlinewould be long-term erosion and landward retreat (DE MOOR,



Figure 4. Evolution of high water level (HW), mean sea level (MSL), and low water level (LW) in Oostende for the periods 1835–2000 and 1930–2000 (Fromvan Cauwenberge, 1999).

1992; DE WOLF et al., 1993), which can probably be attributedto a long-term sediment deficit.

Saltwater Intrusion

At present, monitoring of saltwater intrusion is not carriedout systematically. Only sporadic geo-electrical surveys and

water analyses are made, but usually not with the intentionof monitoring saltwater encroachment.

NATURAL AND ARTIFICIAL COASTAL DEFENCE

The natural defence of the Belgian coast consists of a sandyforeshore, backshore, and dune belt. This natural protection



Table 2. Decadal trends of mean high water (MHW) and mean low waterlevels (MLW) along with the tidal amplitude (MTA) for a period of 100years near Antwerpen (Viaenen, 2000).

Decade MHW (m TAW) MLW (m TAW) MTA (m)

1891–1900 4.68 0.29 4.391901–1910 4.72 0.23 4.491911–1920 4.83 0.24 4.591921–1930 4.85 0.20 4.651931–1940 4.90 0.18 4.721941–1950 4.90 0.17 4.731951–1960 4.96 0.15 4.811961–1970 5.07 0.17 4.901971–1980 5.15 0.01 5.141981–1990 5.24 0.05 5.19

extends along 40% of the coastline with additional artificialdefence structures elsewhere to ensure the preservation ofthe present shape and situation of the coastline. These arti-ficial defences are categorised as hard or soft (MVG, 1993).Soft protection measures that have been applied to the Bel-gian coast include beach nourishment, near-shore nourish-ment, temporary wind screens, sand berms on the back shore,beach scrapings, and planting of osier hedges or marramgrass (MVG, 1993). The Longard tube system was built atLombardsijde and De Haan, accompanied by beach nourish-ment of 40 � 103 m3 over 1.25-km length in 1981 and 12.5 �106 m3 over 4-km length in the period from 1978 to 1980.Other important beach nourishment schemes occurred atBlankenberge (30 � 103 m3 in 1988) and at Knokke (�8 �106 m3 in 1977–1979 and 1 � 106 m3 in 1986). At Knokke, anexperiment with underwater screens to stabilise the beachand prevent cross-shore erosion was performed.

Sea dikes and dikes presently protect coastal cities such asOostende, Zeebrugge, De Haan, and De Panne, etc. againststorm surges and high waves. These dikes are usually com-bined with groynes and shorter beach groynes. Currently, noclear idea or inventory exists about the actual defencestrength of most dikes and sea dikes in Belgium (MVG, 1993),although maintenance and investment costs for coastal de-fence are about €25 million per year (P. DE WOLF, personalcommunication).

Antwerpen and other cities/villages along the Schelde es-tuary are protected against flooding and inundation by dikes.These dikes will shortly be raised to attain the presently re-quired standard level protection against flooding for a stormfrequency of 1/4000 years. Another potential protection mea-sure is the creation of controlled flooding areas, which is nowunder discussion (ZANTING and TEN THIJ, 2001). These areareas that will accommodate the extra water generated dur-ing dangerously high water levels to reduce the threat to ur-ban areas and other valuable land. These measures are partof the SIGMA-plan (plan for the development of steps againstinundations of the river Schelde).

VULNERABILITY OF COASTAL AREAS

Some of the possible implications of sea-level rise for Bel-gium are now considered, with a strong emphasis on the im-plications for groundwater, drainage, and rising water tables.

Flooding and erosion are important concerns. Despite thenatural and artificial defence structures, the coastal plain willalways remain vulnerable to changes in sea level or storm fre-quency. One example is the city of Oostende. At present, onlyan old sea dike protects it against flooding. The beach at Oos-tende is narrow and offers limited protection. The dike wasconstructed for a storm frequency of 250 years. For future up-grade, a storm frequency of 4000 years will be used in thedesign (DE ROECK, 1999). Another part of the flood defenceplan is to widen the shore through beach nourishment (P. DE

WOLF, personal communication). Concrete sea walls counter-act the natural exchange of sediments between foredunes,beach, and offshore (BAETEMAN et al., 1992). Consequently, thesand necessary to maintain the beach profiles under sea-levelrise cannot be delivered by the erosion of the foredune. Thiswill result in accelerated beach erosion, and the underminingof the seawall is not inconceivable if there is not ongoing nour-ishment.

The polder area is protected by dikes and is surroundedand intersected by canals that drain most of the infiltratingfreshwater almost immediately. Sea-level rise would increasethe risk of saltwater seepage into the polder area, especiallyinto the polder areas immediately landward of the narrowdune belts. On the seaward side of the dunes the landwardinfiltration of saltwater will increase because of rising highwater levels. This would result in increased saltwater seep-age into polder areas, an increased amount of drainage water,and an increase of salt content in the drained water. Theincreased infiltration of saltwater on the shore in front of thewide dune belt will result in a larger saltwater lens on theshore, which occurs above the reduced flow of freshwaterfrom the dunes toward the sea (Figure 3). Sea-level rise andany increase of tidal amplitude will result in rising watertables in the dune area, possibly leading to flooding of oldcellars, underground car parks, and other underground ser-vices.

Sea-level rise and any increase in tidal amplitude will re-duce the water extraction capacity in the wide dune belt. Ata constant discharge rate and after a certain rise of the highhigh water level, saltwater will start to flow from the shoreto the extraction wells. Possible measures against this salt-water intrusion are: (i) reducing the extraction rate of thewells close to the seaside, eventually combined with increas-ing the discharge rate of the well close to the polder area and(ii) artificial recharging of the dune area close to the sea,among others. For an optimal design of such adaptation mea-sures, detailed hydrogeological studies will be necessary. Pre-liminary response measures have already been considered(LEBBE et al., 1995). Artificial recharge could be realised withsurface water and/or effluent from a wastewater treatmentplant, including cleansing with ultrafiltration or inverse os-mosis (VAN HOUTTE et al., 1998).

At present, a large part of the surface water of the coastalplain and its surrounding areas flows by gravity to the seathrough locks that are only open at low tides. Because of thislimitation, clay-covered peat areas are regularly flooded dur-ing wetter periods (BAETEMAN et al., 1992). This is particu-larly true for those areas on the right bank of the river IJzer,which has lower dikes than the left bank. Additionally, drain-



age water of the left bank can also be pumped continuouslyinto the sea behind the Nieuwpoort Locks (total dischargerate capacity of 28 m3/s) when the gravitational flow is toosmall. Isolated small, low areas in the polders are alsoequipped with pumps, which lift the water into larger canalswith gravitational flow to the sea (e.g., the reclaimed lakesand the clay-covered peat areas situated near the landwardboundary of the polders). Sea-level rise, tidal amplitude in-creases, and rain intensity increases will all have their influ-ence on the proper management of the drainage system evenif an enlargement of wetland areas is taken into account. Inthese areas a rise of the drainage level is considered (e.g., theUitkerkse polder, 127 ha). Polders requiring improved drain-age in rain intensive periods should be managed to avoid in-creased subsidence.

The Schelde estuary is also an important natural site, in-cluding precious intertidal flats and associated ecosystems. Itis not yet clear if anthropogenic changes in recent years havecaused a threat to these areas and communities. Current sed-iment losses out of the estuary could, along with sea-levelrise, cause drowning of some nature reserve areas, such asthe intertidal flats. Further research on this issue is neces-sary (ZANTING and TEN THIJ, 2001).

CONCLUSION AND FUTURE RESEARCH ISSUES

Although the hinterland is naturally protected by fore-shore, backshore, and dunes, this natural protection existsfor only 40% of the Belgian North Sea coast. The remainderis protected by dikes, sea dikes, groynes, and other (soft) pro-tection measures in an attempt to maintain the present widthof shore and dunes. About €25 million is spent each year onmaintenance of, and investment in, coastal defences. Adap-tation to sea-level rise is necessary to protect the densely pop-ulated coastal area, where economic activity is important anddependent on the present shape of the coastline and on thedrainage level of the coastal plain. Sea-level rise will makemore artificial drainage necessary and will necessitate exten-sive application of protective measures to maintain the coast-line, including the Schelde estuary.

The main morphological features of the coastal plain areartificial or human influenced (cf. VAN KONINGSVELD et al.,2008). Low-lying polders are completely artificial and inter-sected by canals that conduct the water from the higher sur-rounding areas by gravitational flow through the locks to thesea. Sea-level rise and any increase of tidal amplitude willreduce this drainage capacity while, at the same time, theneed for drainage capacity is expected to increase as rain in-tensity increases. Important freshwater resources in thecoastal dunes are also threatened. These qualitative deduc-tions assume that the drainage levels in the polder areas re-main unaltered. Thorough hydrogeological studies are nec-essary to maintain fresh groundwater reserves in the Belgiancoastal plain.

Sea-level rise monitoring shows that high water level hasrisen about an average of 20 mm/decade and mean sea levelan average of 15 mm/decade over the 20th century. Coastalmanagement is focused on preserving the present shoreline.New dike design criteria in Belgium are comparable to Dutch

criteria, which are based on an increasing storm risk and afuture sea-level rise of 6 mm/y (DE ROECK, 1999). Increaseddrainage in the polders will be necessary to remove the extrasaltwater seepage and increased rain intensity. Some studieson the effects of sea-level rise on the Belgian coast have beencarried out (SCHOETERS and VANHAECKE, 1999; VIAENE,2000), but they are based on results from other Europeanstudies, which have then been extrapolated to the Belgiansituation.

This investigation has highlighted the need for future re-search to focus on the quantitative interpretation of the avail-able data, such as numerical modelling of sediment transportand groundwater flow, in order to understand the effect ofsea-level rise on beach erosion, flood risk, fresh and saltwaterdistribution, etc. The collection of additional data on fresh–saltwater distribution, sediment distribution, and the effectof dredging along the shoreline would also be of benefit. Fu-ture research should also consider the IntergovernmentalPanel on Climate Change scenarios of sea-level rise and cli-mate change. This basic data could then be used for planningand reevaluation of coastal management decisions. Besidethe elaboration of physical models, data acquisition, and cal-ibrations, a thorough socio-economic study should be made toassess the vulnerability of the Belgian coast. Lastly, the vul-nerability of the Western Schelde Estuary should be furtherassessed in close collaboration with The Netherlands.

ACKNOWLEDGMENTS

The authors would like to thank three anonymous review-ers. R. Nicholls and A. de la Vega-Leinert are acknowledgedfor their editorial comments. Their suggestions and com-ments enhanced the quality of the paper.

LITERATURE CITED

ALLAERT, G., 1996. De Belgische kust: tekenen van een recente re-vival? Internal note (In Dutch). Department of Areal Survey,Ghent University, Belgium.

BAETEMAN, C.; DE LANNOY, W.; PAEPE, R., and VAN CAUWENBER-GE, C., 1992. Vulnerability of the Belgian coastal lowlands to fu-ture sea-level rise. In: TOOLEY, M.J., and JELGERSMA, S. (eds.),Impacts of Sea-level Rise on European Coastal Lowlands. The In-stitute of British Geographers Special Publications Series 27. Ox-ford, UK: Blackwell, pp. 56–71.

BAETEMAN, C. and DENYS, L., 1997. Holocene shoreline and sea-level data from the Belgian coast. Paleoclimate Research, 21, 49–74.

BEAR, J., 1979. Hydraulics of Groundwater. New York: McGraw-Hill.DE BREUK, W.; DE MOOR, G.; MARECHAL, R., and TAVERNIER, R.,

1974. Depth of the salt-freshwater interface in the phreatic aquiferin the Belgian coastal plain (1963–1973). Proceedings of the 4thSaltwater Intrusion Meeting (Ghent, Belgium). Appendix (map,scale 1/100.000).

DE KRAKER, A.M.J., 1999. A method to assess the impact of hightides, storms and storm surges as vital elements in climatic his-tory the case of stormy weather and dikes in the northern part ofFlanders, 1488 to 1609. In: PFISTER, C., BRADZIL, R., and GLASER,R. (eds.), Climatic Variability in Sixteenth Century Europe and ItsSocial Dimension (reprinted from Climatic Change, 43,[1]). Dor-drecht, The Netherlands: Kluwer Academic Publishers, pp. 287–302.

DE MOOR, G., 1992. A quantitative evaluation of erosive and accre-tional sections along the Belgian coast in the period 1978–1990.



Tijdschrift van de Belgische Vereniging voor AardrijkskundigeStudies (BEVAS), 2, 413–424.

DE ROECK, J., 1999. Advies betreffende de zeedijk te Oostende tussenhet casino en het klein strand. Internal note (In Dutch). Waterwaysand Maritime Affairs Administration, Belgium.

DE WOLF, P.; FRANSNAER, D.; VAN SIELEGHEM, J., and HOUTHUYS,R., 1993. Morphological trends of the Belgian coast shown by 10years of remote-sensing based surveying. In: Proceedings of the 8thSymposium on Coastal and Ocean Management (19–23 July; NewOrleans, Louisiana), pp. 245–257.

DENYS, L. and BAETEMAN, C., 1995. Holocene evolution of relativesea level and local mean high water spring tides in Belgium–a firstassessment. Marine Geology, 124, 1–19.

DENYS, L.; LEBBE, L.; SLIGGERS, L.C.; SPAINK, G.; VAN, S.M., andVERBRUGGEN, C., 1983. Litho- and biostratigraphical study ofquaternary deep marine deposits of the western Belgium coastalplain. Bulletin van de Belgische Vereniging voor Geologie, T 92(2),125–154.

DOMENICO, P.A. and SCHWARTZ, F.W., 1990. Physical and ChemicalHydrogeology. New York: John Wiley and Sons.

DOUGLAS, B.C., 2001. Sea level change in the era of the recordingtide gauge. In: DOUGLAS, B.C., KEARNEY, M.S., and LEATHER-MAN, S.P. (eds.), Sea Level Rise, History and Consequences. NewYork: Academic Press, pp. 37–64.

FREEZE, R.A. and CHERRY, J.A., 1979. Groundwater. EnglewoodCliffs, New Jersey: Prentice-Hall.

LEBBE, L. and PEDE, K., 1986. Salt-freshwater flow underneath theold dunes and low polders influenced by pumpage and drainagein the western Belgian coastal plain. In: Proceedings of the 9thSaltwater Intrusion Meeting. Delft, the Netherlands, pp. 199–220.

LEBBE, L.; TARHOUNI, J.; VAN HOUTTE, E., and DE BREUCK, W.,1995. Results of an artificial recharge test and a double pumpingtest as preliminary studies for optimising water supply in thewestern Belgian coastal plain. Hydrogeology Journal, 5(3), 44–55.

LEBBE, L., 1999. Parameter identification in fresh-saltwater flowbased on borehole resistivities and freshwater head data. Advancesin Water Resources, 22(8), 791–806.

LOUWYE, S. and DECLERCQ, E., 1998. Relative water level changein the intracoastal zone of Belgium and Northern France over thelast 2500 years. Boreas, 27, 162–177.

MOSTAERT, F., 1985. Een vergelijkende studie van de sedimentologieen detailstratigrafie van Eemiaan en Holocene getijdesedimentenin het oostelijk deel van de kustvlakte. Ghent, The Netherlands:Ghent University, Ph.D. thesis (In Dutch).

MOSTAERT, F., 1986. Kreekruggen en komgronden in het Bruse Oud-land. Kritische analyse van de verschillende morfogenetische in-

terpretaties. Critical analysis of different morphogenetic interpre-tations. Werkstukken, XXV (In Dutch). Nationaal centrum voorgeomorfologisch onderzoek (NCGO).

MVG (Ministerie van de Vlaamse Gemeenschap, the Ministry of theFlemish Community), 1993. Kust 2002. Deel 1: De zeewerendefunctie van de kust. Stand van zaken. Preparational report forDepartment of Natural Environment and Infrastructure, Water-ways and Maritime Affairs Administration, Harbour council (InDutch). Ministry of the Flemish Community.

SCHOETERS, K. and VANHAECKE, P., 1999. Kader voor rapporteringvan ‘‘Climate Change’’ effecten in Belgie: uitwerking en toepass-ing. Final report on the effects of Climate Change in Belgium:development and application (In Dutch). Ecolas, N.V. for FederalService for Scientific, Technological and Cultural Affairs, 121p.

TAVERNIER, R.; AMERYCKX, J.; SNACKEN, F., and FARASYN, D.,1970. Kust, Duinen en Polders. In: Atlas van Belgie, Plaat 17, Be-geleidende tekst. Brussels, Belgium: Militair Geografisch Instituut,pp. 32.

TODD, D.K., 1980. Groundwater Hydrology. 2nd edition. New York:John Wiley and Sons.

VAN CAUWENBERGE, C., 1999. Relative Sea-Level Rise Along theBelgian Coast: Analyses and Conclusions With Respect to theHigh Water, the Mean Sea and the Low Water Levels. Report 46.Hydrographical Service (Oostende), Water-ways and Maritime Ad-ministration.

VAN HOUTTE, E.; VERBOUWHEDE, J.; VANLERBERGHE, F.; DEMUN-TER, S., and CABOOTER, J., 1998. Treating different types of rawwater with micro- and ultrafiltration for further desalination us-ing reverse osmosis. Desalination, 117, 49–60.

VAN KONINGSVELD, M.; MULDER, J.P.M.; STIVE, M.J.F.; VAN-DERVALK, L., and VANDERWECK, A.W., 2008. Living with sea-level rise and climate change: a case study of the Netherlands.Journal of Coastal Research, 24, 367–379.

VIAENE, P., 2000. Effecten van een mogelijke klimaatverandering ophet zeespiegelniveau, de rivierafvoer en de frequentie van hoog-waters en stormen. Literature study (In Dutch). Ministry of theFlemish Community, Hydraulic Laboratory and Hydrological Re-search Division.

WARRICK, R.A., 1993. Climate and sea level change: a synthesis. In:WARRICK, R.A., BARROW, E.M., and WIGLEY, T.M.L. (eds.), Cli-mate and Sea-level Change, Observations, Projections and Impli-cations. Cambridge, UK: Cambridge University Press, pp. 424.

ZANTING, H.A. and TEN THIJ, F., 2001. Lange termijn visie Scheldeestuarium. Rapport Technische Schelde commissie, RA/00-45 (InDutch). Delft, The Netherlands: RIKZ, WBL and Delft Hydraulics.

� SAMENVATTING �

Zelfs al is de kustvlakte en het estuarium van de Schelde van groot economisch belang voor Belgie en leven er ca. 0.8 miljoen mensen (op een totale bevolking van10 miljoen), tot nu toe bestaat slechts kwalitatieve beschrijvingen over de effecten van een zeespiegelstijging op de kustvlakte. Er zijn geen kwantitatieve modellenontwikkeld die steunen op werkelijke veldwaarnemingen. Hier worden de fysische eigenschappen van het kustgebied beschreven samen met een kwalitatieveinterpretatie van zijn kwetsbaarheid. De laag-gelegen poldergebieden, die kustmatig aangelegd werden, zullen het meest beınvloed worden door een zeespiegelstijging.Een van de belangrijkste problemen in dit gebied is de waterevacuatie tijdens regenrijke periodes. Verschillende delen van de polders worden op verschillende wijzendroog gehouden. Hun verschillende kwetsbaarheid voor zeespiegelstijging en toename van regenidensiteit wordt ingeschat. Het verband tussen de drainageniveau’sen verziltingsgevaar wordt eveneens beschreven. Ook de zoetwaterlenzen, ontwikkeld onder de duinen en geexploiteerd voor drinkwater, worden door een zeespie-gelstijging bedreigd. De kustverdedigingsstructuren en hun efficientie worden besproken. De effecten van een zeespiegelstijging voor de haven van Antwerpen envoor het Schelde estuarium zijn ver van duidelijk. De resultaten van het opvolgen van de zeespiegelstijging tijdens de 20ste eeuw, van het waarnemen van de windenen golven en de evolutie van erosie en accretie worden eveneens beschreven. Toekomstig onderzoek zou zich vooral moeten richten op de quantitatieve interpretatievan gegevens om zeespiegelstijgingseffecten in te schatten op stranderosie, overstromingsrisico’s en zoet-zout-grondwaterdistributies Naast deze fysische modellenzou een grondige socio-economische studie moeten gemaakt worden, resulterend in een inschatting van de kwestbaarheid van het Belgische kustgebied en van hetScheldegebied.

Documents

Potential Implications of Sea-Level Rise for Belgium